Sampling and Observational Studies

Observational studies

Sample a group of people and watch/observe their behavior (no control/watch and observe from the outside)

Experiments

Recruit a group of people as assign them to a treatment (assign to certain behavior)

Effective sampling

Sample must represent the population

Sample must not be biased

Key to effective sampling

Randomization

Simple random sample

Randomly sample individuals so that all samples of same size have equal chance of being sampled

Cluster sample

Put population into groups where each group is representative of the population then randomly sample a few groups and include all individuals in selected groups (All of some groups)

Stratified sample

Put population into groups where each group is similar within the groups then randomly sample some individuals in each group (some of all groups)

Multi-stage sample

Successively sample groups until you sample individuals